178 PHYSIOLOGIC GENETICS 



does the hypothesis of a great many factors with contributions in geometric series 

 (40 per cent: 24 percent: 14.4 per cent: 8.6 percent, and so forth). 1448 Results in third 

 backcrosses best fit an intermediate hypothesis somewhat similar to the last. 



It is assumed that the effects are additive on the physiologic scale. However, 

 the factors were not isolated. It may be suspected that, if they had been isolated, a 

 more complex situation would have been found with specific interactions as in the 

 cases of color and hair direction. As indicated above, the hypothesis of two thresholds 

 permits considerable elasticity with respect to the effects on the underlying scale. 



Similar small toes have been restored in a wholly different way, in this case often 

 associated with thumbs and great toes resembling those of other rodents. 1413 A single 

 mutant individual showed those characters imperfectly developed. From thousands of 

 descendants it appears that a dominant gene (Px) with variable penetrance was respon- 

 sible. In the stock of origin, about 82 per cent showed one or more atavistic digits. 

 After certain crosses, penetrance fell to 20 per cent or less. Crosses with strain D 

 revealed interesting interaction effects. It was not surprising that penetrance of the 

 small toe in Pxpx rose from 62 to 100 per cent by combining these two heredities which 

 favor it. More interesting is the fact that penetrance of the thumb rose from 74 to 

 100 per cent, although thumbs were wholly absent in strain D. Similarly, the pene- 

 trance of the great toe rose from 2 to 18 per cent after one backcross to D and to 55 per 

 cent after two backcrosses. 



The homozygote PxPx turned out to be lethal and grossly abnormal. It was found 

 by Scott 1184 that about 92 per cent died and were absorbed at about the twenty- 

 eighth day of gestation. Those that reached birth 1185 had short legs, rotated hind legs 

 without tibiae, and broad paddle-shaped feet with 7 to 12 digits each. The animals 

 were always microphthalmic and either hydrocephalic or with bi'ains protruding from 

 the skull, usually harelipped, and grossly abnormal in most of their internal anatomy. 

 The defect was manifest at about 18 days of gestation in limb buds of double width and 

 overgrown midbrains and hindbrains. This constitutes an extreme case of pleiotropy. 



The atavistic return of digits, lost in the evolution of the whole family Caviidae, 

 raises a question on the nature of homology. Homology is usually treated as an all-or- 

 none matter: particular organs of two species either are or are not homologous. The 

 language used in discussing evolution of organs often seems to imply preformation of 

 an extreme sort, a heredity evolving separately for each part. 



It is obvious, however, from the genetics of morphologic characters in all organisms 

 studied, that replicative homologs develop to a large extent under the same systems of 

 genes. Thus genotype RRMM has similar effects on hair direction and the underlying 

 pattern of growth of the skin, not only on right and left hind feet but on hind feet and 

 fore feet (with some difference in threshold) and on the separate digits on these feet 

 (a pattern of partings along the sides and across the upper side of these digits shaped 

 like a letter H). The situation is similar with the genes that tend to restore the 

 pentadactyl foot. It is supposed that all of the genes under which any part of the ances- 

 tral pentadactyl foot developed were so deeply involved in the development of other 



